INDEX

Explanations

names related to military or political figures

oai_token-act-pair · gpt-3.5-turbo

references to specific individuals, particularly in the context of military or government-related discussions

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.hook_resid_pre

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.5.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ting

-0.80

ty

-0.80

ted

-0.77

vice

-0.75

theless

-0.69

ties

-0.68

Ķ

-0.66

 Spielberg

-0.64

istic

-0.64

cision

-0.63

POSITIVE LOGITS

aurus

1.09

pread

0.98

hift

0.97

CRIP

0.94

hips

0.92

aeda

0.90

chwitz

0.90

creen

0.89

andra

0.88

cale

0.88

Activations Density 0.052%

names related to military or political figures

references to specific individuals, particularly in the context of military or government-related discussions

No Comments

No Known Activations